[go: up one dir, main page]

CN112614142A - Cell weak label manufacturing method and system based on multi-channel image fusion - Google Patents

Cell weak label manufacturing method and system based on multi-channel image fusion Download PDF

Info

Publication number
CN112614142A
CN112614142A CN202011560828.4A CN202011560828A CN112614142A CN 112614142 A CN112614142 A CN 112614142A CN 202011560828 A CN202011560828 A CN 202011560828A CN 112614142 A CN112614142 A CN 112614142A
Authority
CN
China
Prior art keywords
cell
image
subgraph
stacked
stained
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202011560828.4A
Other languages
Chinese (zh)
Other versions
CN112614142B (en
Inventor
温廷羲
潘婷
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Huaqiao University
Original Assignee
Huaqiao University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Huaqiao University filed Critical Huaqiao University
Priority to CN202011560828.4A priority Critical patent/CN112614142B/en
Publication of CN112614142A publication Critical patent/CN112614142A/en
Application granted granted Critical
Publication of CN112614142B publication Critical patent/CN112614142B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/11—Region-based segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/23—Clustering techniques
    • G06F18/232—Non-hierarchical techniques
    • G06F18/2321—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions
    • G06F18/23213—Non-hierarchical techniques using statistics or function optimisation, e.g. modelling of probability density functions with fixed number of clusters, e.g. K-means clustering
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F18/00—Pattern recognition
    • G06F18/20—Analysing
    • G06F18/25—Fusion techniques
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00—Machine learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/136—Segmentation; Edge detection involving thresholding
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T7/00—Image analysis
    • G06T7/10—Segmentation; Edge detection
    • G06T7/187—Segmentation; Edge detection involving region growing; involving region merging; involving connected component labelling
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V10/00—Arrangements for image or video recognition or understanding
    • G06V10/20—Image preprocessing
    • G06V10/28—Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
    • G06V20/00—Scenes; Scene-specific elements
    • G06V20/60—Type of objects
    • G06V20/69—Microscopic objects, e.g. biological cells or cellular parts
    • G06V20/695—Preprocessing, e.g. image segmentation
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/10—Image acquisition modality
    • G06T2207/10056—Microscopic image
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20081—Training; Learning
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20092—Interactive image processing based on input by user
    • G06T2207/20101—Interactive definition of point of interest, landmark or seed
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T2207/00—Indexing scheme for image analysis or image enhancement
    • G06T2207/20—Special algorithmic details
    • G06T2207/20112—Image segmentation details
    • G06T2207/20152—Watershed segmentation
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02A—TECHNOLOGIES FOR ADAPTATION TO CLIMATE CHANGE
    • Y02A90/00—Technologies having an indirect contribution to adaptation to climate change
    • Y02A90/10—Information and communication technologies [ICT] supporting adaptation to climate change, e.g. for weather forecasting or climate simulation

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Data Mining & Analysis (AREA)
  • General Engineering & Computer Science (AREA)
  • Evolutionary Computation (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Biology (AREA)
  • Software Systems (AREA)
  • Bioinformatics & Cheminformatics (AREA)
  • Bioinformatics & Computational Biology (AREA)
  • Multimedia (AREA)
  • Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Medical Informatics (AREA)
  • Probability & Statistics with Applications (AREA)
  • Computing Systems (AREA)
  • Mathematical Physics (AREA)
  • Image Analysis (AREA)
  • Image Processing (AREA)
  • Investigating Or Analysing Biological Materials (AREA)

Abstract

The invention provides a method and a system for manufacturing a cell weak label based on multi-channel image fusion, which belongs to the technical field of machine learning, and the method comprises the following steps: step S10, obtaining a staining cell nucleus image and a staining cytoplasm image; step S20, performing binarization processing on the stained cell nucleus image to obtain a connected cell subgraph; step S30, setting an area threshold, and segmenting stacked images and non-stacked images from the connected cell subgraph based on the area threshold; step S40, segmenting cytoplasm subgraphs with the same position and size from the dyed cytoplasm images based on the stacked images, and segmenting the stacked images based on the cytoplasm subgraphs by using a watershed algorithm to obtain a first segmentation result; step S50, segmenting the non-stacked image by using a watershed algorithm to obtain a second segmentation result; and step S60, combining the first segmentation result and the second segmentation result to obtain a weak label of the stained cell nucleus image. The invention has the advantages that: the confidence of the weak label of the cell is greatly improved.

Description

Cell weak label manufacturing method and system based on multi-channel image fusion
Technical Field
The invention relates to the technical field of machine learning, in particular to a method and a system for manufacturing a cell weak label based on multi-channel image fusion.
Background
Machine learning requires labeling of training sample data in advance, and currently, expensive and time-consuming manual labeling has become an important bottleneck for high-quality machine learning model research and application. In machine learning, the model fitting training using labeled data with high confidence (e.g., manual labeling) is supervised learning, and the model fitting training using labeled data with relatively low confidence (e.g., weak labeling) is weakly supervised learning. In example segmentation of cells, model supervised learning by weak labeling is an important way to reduce cost to drive research.
Common labeling methods for weak supervised learning in semantic segmentation include Scribes, Box, Point and the like, but these labeling methods can only roughly label the position of a target object in an image, and have no clear edge information, and the quality of generating a pixel-level weak label directly affects the performance of a segmentation model.
Therefore, how to provide a method and a system for manufacturing a cell weak label based on multi-channel image fusion to improve the confidence of the cell weak label becomes a problem to be solved urgently.
Disclosure of Invention
The invention aims to provide a method and a system for manufacturing a cell weak label based on multi-channel image fusion, so as to improve the confidence of the cell weak label.
In a first aspect, the invention provides a method for manufacturing a cell weak label based on multi-channel image fusion, which comprises the following steps:
step S10, obtaining a stained cell nucleus image and a stained cytoplasm image of the same cell colony;
step S20, performing binarization processing on the stained cell nucleus image to obtain a connected cell subgraph in the stained cell nucleus image after binarization processing;
step S30, setting an area threshold, and segmenting stacked images and non-stacked images from the connected cell subgraph based on the area threshold;
step S40, segmenting cytoplasm subgraphs with the same position and size from the dyed cytoplasm image based on the stacked image, and segmenting the stacked image based on the cytoplasm subgraphs by using a watershed algorithm to obtain a first segmentation result;
step S50, segmenting the non-stacked image by using a watershed algorithm to obtain a second segmentation result;
and step S60, combining the first segmentation result and the second segmentation result to obtain a weak label of the stained cell nucleus image.
Further, the step S30 specifically includes:
step S31, setting an area threshold, sequentially judging whether the area of each connected cell subgraph is larger than the area threshold, if so, indicating that a cell stacking region exists, and entering step S32; if not, indicating that no cell stacking area exists, and ending the process;
and step S32, selecting the connected cell subgraph by using a circumscribed rectangle frame, determining a cell stacking area of the connected cell subgraph by using a centroid, and further segmenting the connected cell subgraph by using the cell stacking area to obtain a stacked image and a non-stacked image.
Further, in the step S40, the watershed algorithm is based on H-minima adaptive flag control.
In a second aspect, the invention provides a cell weak label making system based on multi-channel image fusion, which comprises the following modules:
the cell image acquisition module is used for acquiring a staining cell nucleus image and a staining cytoplasm image of the same cell community;
the connected cell subgraph acquisition module is used for carrying out binarization processing on the stained cell nucleus image and acquiring a connected cell subgraph in the stained cell nucleus image after the binarization processing;
the connected cell subgraph segmentation module is used for setting an area threshold value and segmenting a stacked image and a non-stacked image from the connected cell subgraph based on the area threshold value;
the stacked image segmentation module is used for segmenting cytoplasm subgraphs with the same position and size from the dyed cytoplasm image based on the stacked image, and segmenting the stacked image based on the cytoplasm subgraphs by using a watershed algorithm to obtain a first segmentation result;
the non-stacked image segmentation module is used for segmenting the non-stacked image by utilizing a watershed algorithm to obtain a second segmentation result;
and the weak label generating module is used for combining the first segmentation result and the second segmentation result to obtain a weak label of the stained cell nucleus image.
Further, the connected cell subgraph segmentation module specifically comprises:
the area comparison unit is used for setting an area threshold value, sequentially judging whether the area of each connected cell subgraph is larger than the area threshold value, if so, indicating that a cell stacking region exists, and entering the cell stacking region determination unit; if not, indicating that no cell stacking area exists, and ending the process;
and the cell stacking area determining unit is used for selecting the connected cell subgraph by using a circumscribed rectangle frame, determining the cell stacking area of the connected cell subgraph by using a centroid, and further segmenting the connected cell subgraph by using the cell stacking area to obtain a stacked image and a non-stacked image.
Further, in the stacked image segmentation module, the watershed algorithm is based on an H-minima adaptive marker control.
The invention has the advantages that:
through fusing channel information of the staining cell nucleus image and the staining cytoplasm image and combining a watershed algorithm, pixel-level segmentation is carried out on the communicated cell subgraph in the staining cell nucleus image one by one to serve as a weak label of the staining cell nucleus image.
Drawings
The invention will be further described with reference to the following examples with reference to the accompanying drawings.
FIG. 1 is a flow chart of a cell weak label manufacturing method based on multi-channel image information fusion.
FIG. 2 is a schematic structural diagram of a cell weak label making system based on multi-channel image information fusion.
FIG. 3 is a schematic diagram of the cell weak label extraction process of the present invention.
Detailed Description
Referring to fig. 1 to fig. 3, a preferred embodiment of a method for manufacturing a cell weak label based on multi-channel image information fusion according to the present invention includes the following steps:
step S10, obtaining a stained cell nucleus image and a stained cytoplasm image of the same cell colony;
step S20, performing binarization processing on the stained cell nucleus image to obtain a connected cell subgraph in the stained cell nucleus image after binarization processing;
step S30, setting an area threshold, and segmenting stacked images and non-stacked images from the connected cell subgraph based on the area threshold;
step S40, segmenting cytoplasm subgraphs with the same position and size from the dyed cytoplasm image based on the stacked image, and segmenting the stacked image based on the cytoplasm subgraphs by using a watershed algorithm to obtain a first segmentation result; before segmentation, the cytoplasm subgraphs and the stacked images are tiled and spliced;
step S50, segmenting the non-stacked image by using a watershed algorithm to obtain a second segmentation result;
and step S60, combining the first segmentation result and the second segmentation result to obtain a weak label of the stained cell nucleus image.
The step S30 specifically includes:
step S31, setting an area threshold, sequentially judging whether the area of each connected cell subgraph is larger than the area threshold, if so, indicating that a cell stacking region exists, and entering step S32; if not, indicating that no cell stacking area exists, and ending the process;
and step S32, selecting the connected cell subgraph by using a circumscribed rectangle frame, determining a cell stacking area of the connected cell subgraph by using a centroid, and further segmenting the connected cell subgraph by using the cell stacking area to obtain a stacked image and a non-stacked image.
In step S40, the watershed algorithm is based on H-minima adaptive token control.
The watershed algorithm segmentation process is as follows:
firstly, carrying out image preprocessing on the images obtained by tiling and splicing the cytoplasmic subgraph and the stacked images based on morphology, and setting S as an obtained set of N cell stacked regions (S is S)j(j∈[1,...,N]) S is used as the input of a watershed algorithm, and the output is a segmentation set of a cell stacking region
Figure BDA0002860381150000051
And defining a new roundness measurement index FuzzyR, taking the average roundness maximum value of each cell in the candidate segmentation result corresponding to each h value as a clustering result evaluation target, and converting the self-adaptive selection problem of the optimal h value of the cell stacking area into a clustering number optimization problem for improving a K-means clustering segmentation algorithm. Once the optimal h value of the stacking area is determined, the candidate segmentation result corresponding to the h value is the optimal segmentation result.
Firstly, initializing j to 1, performing distance transformation on any cell stacking region, and performing initial segmentation by using a watershed algorithm to obtain m sub-segmentation regions, wherein Cj＝{Cj(i) I 1.. m } is SjAnd obtaining SjA Region Adjacency Graph (RAG) of the middle m sub-partitioned regions. At the same time, extracting SjEstablishing a clustering sample set F by using the characteristics of the mass center abscissa, the ordinate, the area gray mean value, the area gray variance and the like of the m sub-areas, wherein the weight coefficients of the four characteristics are 0.25, 0.3 and 0.2 respectively; if m is 1, SjI.e. the final segmentationThe result is; otherwise, SjIs a cell stacking region. When S isjIn the case of the cell stacking region, h is set to 0,
Figure BDA0002860381150000056
opt_cluster＝0。
if N is presentj(h)≥1&&Nj(h)≤NjWhen (h-1) is present, the number of atoms is Nj(h) For the number of clusters of the clustering algorithm, corej(h) The candidate seed region centroid point is an initial clustering center point, and an RAG-based improved K-means clustering algorithm is adopted to generate a cluster containing Nj(h) Candidate segmentation result SR of individual cellj(h) In that respect Otherwise, repeating the previous step after h is h + delta h. Finally calculating candidate segmentation result SRj(h) In which contains Nj(h) Average circularity of individual cells vfr (h). The average circularity vfr (h) is calculated as follows:
Figure BDA0002860381150000052
denotes the value N after h is selectedj(h) And carrying out clustering region combination for the clustering number to obtain the average roundness of each cell in the candidate segmentation result. Wherein the roundness measure index fuzzzyr is defined as follows:
Figure BDA0002860381150000053
if opt _ cluster < VFR (h), then opt-cluster ═ VFR (h),
Figure BDA0002860381150000054
Figure BDA0002860381150000055
is SjAnd finally, segmenting the result. Let j equal j +1 and repeat all the above processes until j > n. H meeting the optimization target opt _ cluster is a connected region SjAdaptive adaptation
Figure BDA0002860381150000061
The optimal segmentation result is
Figure BDA0002860381150000062
The invention relates to a preferred embodiment of a cell weak label manufacturing system based on multi-channel image information fusion, which comprises the following modules:
the cell image acquisition module is used for acquiring a staining cell nucleus image and a staining cytoplasm image of the same cell community;
the connected cell subgraph acquisition module is used for carrying out binarization processing on the stained cell nucleus image and acquiring a connected cell subgraph in the stained cell nucleus image after the binarization processing;
the connected cell subgraph segmentation module is used for setting an area threshold value and segmenting a stacked image and a non-stacked image from the connected cell subgraph based on the area threshold value;
the stacked image segmentation module is used for segmenting cytoplasm subgraphs with the same position and size from the dyed cytoplasm image based on the stacked image, and segmenting the stacked image based on the cytoplasm subgraphs by using a watershed algorithm to obtain a first segmentation result; before segmentation, the cytoplasm subgraphs and the stacked images are tiled and spliced;
the non-stacked image segmentation module is used for segmenting the non-stacked image by utilizing a watershed algorithm to obtain a second segmentation result;
and the weak label generating module is used for combining the first segmentation result and the second segmentation result to obtain a weak label of the stained cell nucleus image.
The connected cell subgraph segmentation module specifically comprises:
the area comparison unit is used for setting an area threshold value, sequentially judging whether the area of each connected cell subgraph is larger than the area threshold value, if so, indicating that a cell stacking region exists, and entering the cell stacking region determination unit; if not, indicating that no cell stacking area exists, and ending the process;
and the cell stacking area determining unit is used for selecting the connected cell subgraph by using a circumscribed rectangle frame, determining the cell stacking area of the connected cell subgraph by using a centroid, and further segmenting the connected cell subgraph by using the cell stacking area to obtain a stacked image and a non-stacked image.
In the stacked image segmentation module, the watershed algorithm is based on H-minima adaptive landmark control.
The watershed algorithm segmentation process is as follows:
firstly, carrying out image preprocessing on the images obtained by tiling and splicing the cytoplasmic subgraph and the stacked images based on morphology, and setting S as an obtained set of N cell stacked regions (S is S)j(j∈[1,...,N]) S is used as the input of a watershed algorithm, and the output is a segmentation set of a cell stacking region
Figure BDA0002860381150000071
And defining a new roundness measurement index FuzzyR, taking the average roundness maximum value of each cell in the candidate segmentation result corresponding to each h value as a clustering result evaluation target, and converting the self-adaptive selection problem of the optimal h value of the cell stacking area into a clustering number optimization problem for improving a K-means clustering segmentation algorithm. Once the optimal h value of the stacking area is determined, the candidate segmentation result corresponding to the h value is the optimal segmentation result.
Firstly, initializing j to 1, performing distance transformation on any cell stacking region, and performing initial segmentation by using a watershed algorithm to obtain m sub-segmentation regions, wherein Cj＝{Cj(i) I 1.. m } is SjAnd obtaining SjA Region Adjacency Graph (RAG) of the middle m sub-partitioned regions. At the same time, extracting SjEstablishing a clustering sample set F by using the characteristics of the mass center abscissa, the ordinate, the area gray mean value, the area gray variance and the like of the m sub-areas, wherein the weight coefficients of the four characteristics are 0.25, 0.3 and 0.2 respectively; if m is 1, SjNamely the final segmentation result; otherwise, SjIs a cell stacking region. When S isjIn the case of the cell stacking region, h is set to 0,
Figure BDA0002860381150000072
opt_cluster＝0。
if N is presentj(h)≥1&&Nj(h)≤NjWhen (h-1) is present, the number of atoms is Nj(h) For the number of clusters of the clustering algorithm, corej(h) The candidate seed region centroid point is an initial clustering center point, and an RAG-based improved K-means clustering algorithm is adopted to generate a cluster containing Nj(h) Candidate segmentation result SR of individual cellj(h) In that respect Otherwise, repeating the previous step after h is h + delta h. Finally calculating candidate segmentation result SRj(h) In which contains Nj(h) Average circularity of individual cells vfr (h). The average circularity vfr (h) is calculated as follows:
Figure BDA0002860381150000073
denotes the value N after h is selectedj(h) And carrying out clustering region combination for the clustering number to obtain the average roundness of each cell in the candidate segmentation result. Wherein the roundness measure index fuzzzyr is defined as follows:
Figure BDA0002860381150000074
if opt _ cluster < VFR (h), then opt-cluster ═ VFR (h),
Figure BDA0002860381150000081
Figure BDA0002860381150000082
is SjAnd finally, segmenting the result. Let j equal j +1 and repeat all the above processes until j > n. H meeting the optimization target opt _ cluster is a connected region SjAdaptive adaptation
Figure BDA0002860381150000083
The optimal segmentation result is
Figure BDA0002860381150000084
In summary, the invention has the advantages that:
through fusing channel information of the staining cell nucleus image and the staining cytoplasm image and combining a watershed algorithm, pixel-level segmentation is carried out on the communicated cell subgraph in the staining cell nucleus image one by one to serve as a weak label of the staining cell nucleus image.
Although specific embodiments of the invention have been described above, it will be understood by those skilled in the art that the specific embodiments described are illustrative only and are not limiting upon the scope of the invention, and that equivalent modifications and variations can be made by those skilled in the art without departing from the spirit of the invention, which is to be limited only by the appended claims.

Claims (6)

1.一种基于多通道图像融合的细胞弱标签制作方法，其特征在于：包括如下步骤：1. a cell weak label making method based on multi-channel image fusion, is characterized in that: comprise the steps: 步骤S10、获取相同细胞群落的染色细胞核图像以及染色细胞质图像；Step S10, obtaining a stained nucleus image and a stained cytoplasm image of the same cell population; 步骤S20、对所述染色细胞核图像进行二值化处理，获取二值化处理后的所述染色细胞核图像中的连通细胞子图；Step S20, performing a binarization process on the stained cell nucleus image to obtain a connected cell subgraph in the stained cell nucleus image after the binarization process; 步骤S30、设定一面积阈值，基于所述面积阈值从所述连通细胞子图分割出堆叠图像以及非堆叠图像；Step S30, setting an area threshold, and segmenting a stacked image and a non-stacked image from the connected cell subgraph based on the area threshold; 步骤S40、基于所述堆叠图像从染色细胞质图像中分割出相同位置及大小的细胞质子图，利用分水岭算法基于所述细胞质子图对堆叠图像进行分割，得到第一分割结果；Step S40, segmenting the cytoplasmic image with the same position and size from the stained cytoplasm image based on the stacked image, and using a watershed algorithm to segment the stacked image based on the cytoplasmic image to obtain a first segmentation result; 步骤S50、利用分水岭算法对所述非堆叠图像进行分割，得到第二分割结果；Step S50, using a watershed algorithm to segment the non-stacked images to obtain a second segmentation result; 步骤S60、合并所述第一分割结果与第二分割结果得到染色细胞核图像的弱标签。Step S60, combining the first segmentation result and the second segmentation result to obtain a weak label of the stained cell nucleus image. 2.如权利要求1所述的一种基于多通道图像融合的细胞弱标签制作方法，其特征在于：所述步骤S30具体包括：2. The method for making weak cell labels based on multi-channel image fusion according to claim 1, wherein the step S30 specifically comprises: 步骤S31、设定一面积阈值，依次判断各所述连通细胞子图的面积是否大于面积阈值，若是，说明存在细胞堆叠区域，进入步骤S32；若否，说明不存在细胞堆叠区域，并结束流程；Step S31, setting an area threshold, and sequentially judging whether the area of each of the connected cell subgraphs is greater than the area threshold, if yes, it means that there is a cell stacking area, and then go to step S32; if not, it means that there is no cell stacking area, and the process ends ; 步骤S32、利用外接矩形框选所述连通细胞子图，使用形心确定所述连通细胞子图的细胞堆叠区域，进而利用所述细胞堆叠区域对连通细胞子图进行分割，得到堆叠图像以及非堆叠图像。Step S32, use the circumscribed rectangle to frame the connected cell subgraph, use the centroid to determine the cell stacking area of the connected cell subgraph, and then use the cell stacking area to segment the connected cell subgraph to obtain a stacked image and a non-connected cell subgraph. stack images. 3.如权利要求1所述的一种基于多通道图像融合的细胞弱标签制作方法，其特征在于：所述步骤S40中，所述分水岭算法基于H-minima自适应标志控制。3 . The method for making weak cell labels based on multi-channel image fusion according to claim 1 , wherein in the step S40 , the watershed algorithm is based on H-minima adaptive label control. 4 . 4.一种基于多通道图像融合的细胞弱标签制作系统，其特征在于：包括如下模块：4. A cell weak label making system based on multi-channel image fusion is characterized in that: comprising the following modules: 细胞图像获取模块，用于获取相同细胞群落的染色细胞核图像以及染色细胞质图像；The cell image acquisition module is used to acquire the stained nucleus images and stained cytoplasm images of the same cell population; 连通细胞子图获取模块，用于对所述染色细胞核图像进行二值化处理，获取二值化处理后的所述染色细胞核图像中的连通细胞子图；a connected cell subgraph obtaining module, configured to perform binarization processing on the stained cell nucleus image, and obtain the connected cell subgraph in the stained cell nucleus image after the binarization process; 连通细胞子图分割模块，用于设定一面积阈值，基于所述面积阈值从所述连通细胞子图分割出堆叠图像以及非堆叠图像；a connected cell subgraph segmentation module, configured to set an area threshold, and segment a stacked image and a non-stacked image from the connected cell subgraph based on the area threshold; 堆叠图像分割模块，用于基于所述堆叠图像从染色细胞质图像中分割出相同位置及大小的细胞质子图，利用分水岭算法基于所述细胞质子图对堆叠图像进行分割，得到第一分割结果；a stacked image segmentation module, configured to segment the cytoplasmic image of the same position and size from the stained cytoplasm image based on the stacked image, and use a watershed algorithm to segment the stacked image based on the cytoplasmic image to obtain a first segmentation result; 非堆叠图像分割模块，用于利用分水岭算法对所述非堆叠图像进行分割，得到第二分割结果；a non-stacked image segmentation module, used for segmenting the non-stacked images by using a watershed algorithm to obtain a second segmentation result; 弱标签生成模块，用于合并所述第一分割结果与第二分割结果得到染色细胞核图像的弱标签。The weak label generation module is used for combining the first segmentation result and the second segmentation result to obtain the weak label of the stained cell nucleus image. 5.如权利要求4所述的一种基于多通道图像融合的细胞弱标签制作系统，其特征在于：所述连通细胞子图分割模块具体包括：5. A cell weak label making system based on multi-channel image fusion as claimed in claim 4, characterized in that: the connected cell subgraph segmentation module specifically comprises: 面积比对单元，用于设定一面积阈值，依次判断各所述连通细胞子图的面积是否大于面积阈值，若是，说明存在细胞堆叠区域，进入细胞堆叠区域确定单元；若否，说明不存在细胞堆叠区域，并结束流程；The area comparison unit is used to set an area threshold, and sequentially determine whether the area of each of the connected cell subgraphs is greater than the area threshold. cell stack area, and end the process; 细胞堆叠区域确定单元，用于利用外接矩形框选所述连通细胞子图，使用形心确定所述连通细胞子图的细胞堆叠区域，进而利用所述细胞堆叠区域对连通细胞子图进行分割，得到堆叠图像以及非堆叠图像。a cell stacking area determination unit, configured to use a circumscribed rectangle to frame the connected cell subgraph, use the centroid to determine the cell stacking area of the connected cell subgraph, and then use the cell stacking area to segment the connected cell subgraph, Get stacked images as well as non-stacked images. 6.如权利要求4所述的一种基于多通道图像融合的细胞弱标签制作系统，其特征在于：所述堆叠图像分割模块中，所述分水岭算法基于H-minima自适应标志控制。6 . The cell weak labeling system based on multi-channel image fusion according to claim 4 , wherein in the stacked image segmentation module, the watershed algorithm is based on H-minima adaptive marker control. 7 .
CN202011560828.4A 2020-12-25 2020-12-25 A method and system for making weak cell labels based on multi-channel image fusion Active CN112614142B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011560828.4A CN112614142B (en) 2020-12-25 2020-12-25 A method and system for making weak cell labels based on multi-channel image fusion

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011560828.4A CN112614142B (en) 2020-12-25 2020-12-25 A method and system for making weak cell labels based on multi-channel image fusion

Publications (2)

Publication Number Publication Date
CN112614142A true CN112614142A (en) 2021-04-06
CN112614142B CN112614142B (en) 2023-05-30

Family

ID=75245022

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011560828.4A Active CN112614142B (en) 2020-12-25 2020-12-25 A method and system for making weak cell labels based on multi-channel image fusion

Country Status (1)

Country Link
CN (1) CN112614142B (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114359325A (en) * 2021-12-31 2022-04-15 南京理工大学智能计算成像研究院有限公司 An accurate segmentation method of cell images based on digital holographic imaging technology

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104392460A (en) * 2014-12-12 2015-03-04 山东大学 Adherent white blood cell segmentation method based on nucleus-marked watershed transformation
CN110223305A (en) * 2019-06-12 2019-09-10 志诺维思（北京）基因科技有限公司 Cell segmentation method, apparatus and readable storage medium storing program for executing
CN110400318A (en) * 2019-06-14 2019-11-01 广东省人民医院(广东省医学科学院) A method and system for segmenting CD4-positive cell nuclei in immunohistochemical pathological images
CN110415255A (en) * 2019-06-14 2019-11-05 广东省人民医院(广东省医学科学院) A method and system for segmenting CD3-positive cell nuclei in immunohistochemical pathological images
CN110458835A (en) * 2019-08-16 2019-11-15 腾讯科技（深圳）有限公司 A kind of image processing method, device, equipment and medium
CN111210447A (en) * 2020-01-10 2020-05-29 福建师范大学 A method and terminal for hierarchical segmentation of hematoxylin-eosin stained pathological images
WO2020252665A1 (en) * 2019-06-18 2020-12-24 香港理工大学 Method and system for segmenting overlapping cytoplasms in medical image

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104392460A (en) * 2014-12-12 2015-03-04 山东大学 Adherent white blood cell segmentation method based on nucleus-marked watershed transformation
CN110223305A (en) * 2019-06-12 2019-09-10 志诺维思（北京）基因科技有限公司 Cell segmentation method, apparatus and readable storage medium storing program for executing
CN110400318A (en) * 2019-06-14 2019-11-01 广东省人民医院(广东省医学科学院) A method and system for segmenting CD4-positive cell nuclei in immunohistochemical pathological images
CN110415255A (en) * 2019-06-14 2019-11-05 广东省人民医院(广东省医学科学院) A method and system for segmenting CD3-positive cell nuclei in immunohistochemical pathological images
WO2020252665A1 (en) * 2019-06-18 2020-12-24 香港理工大学 Method and system for segmenting overlapping cytoplasms in medical image
CN110458835A (en) * 2019-08-16 2019-11-15 腾讯科技（深圳）有限公司 A kind of image processing method, device, equipment and medium
CN111210447A (en) * 2020-01-10 2020-05-29 福建师范大学 A method and terminal for hierarchical segmentation of hematoxylin-eosin stained pathological images

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
丛培盛;孙建忠;: "分水岭算法分割显微图像中重叠细胞", 中国图象图形学报 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114359325A (en) * 2021-12-31 2022-04-15 南京理工大学智能计算成像研究院有限公司 An accurate segmentation method of cell images based on digital holographic imaging technology

Also Published As

Publication number Publication date
CN112614142B (en) 2023-05-30

Similar Documents

Publication Publication Date Title
CN108364288B (en) Segmentation method and device for breast cancer pathological image
WO2023083059A1 (en) Road surface defect detection method and apparatus, and electronic device and readable storage medium
CN102982539B (en) Characteristic self-adaption image common segmentation method based on image complexity
CN105303169B (en) A kind of cell division identification method and its identification device based on slow feature
CN111738295B (en) Image segmentation method and storage medium
CN109584253B (en) Oil abrasive particle image segmentation method
CN108364032A (en) A kind of cervical cancer cell picture recognition algorithm based on convolutional neural networks
CN117876401B (en) Cervix liquid-based lamellar cell image segmentation method based on SAM segmentation model
CN118314352B (en) Crop remote sensing image segmentation method based on patch-level classification labels
CN114492619A (en) A method and device for constructing point cloud data set based on statistics and concavo-convexity
CN116681657B (en) Asphalt pavement disease detection method based on improved YOLOv7 model
CN116883341A (en) Liver tumor CT image automatic segmentation method based on deep learning
CN108664968B (en) An Unsupervised Text Localization Method Based on Text Selection Model
CN112614142A (en) Cell weak label manufacturing method and system based on multi-channel image fusion
CN115601307A (en) Automatic cell detection method
CN112215303B (en) Image understanding method and system based on self-learning attribute
CN116310293B (en) Method for detecting target of generating high-quality candidate frame based on weak supervised learning
CN114743042B (en) Longjing tea quality identification method based on depth characteristics and TrAdaBoost
CN110309727B (en) Building identification model establishing method, building identification method and building identification device
CN116524297B (en) Weak supervision learning training method based on expert feedback
CN114170218B (en) Chromosome image instance label generation method and system
CN116503591A (en) Image Segmentation Method Based on Weakly Supervised Neural Networks with Few Samples
CN110826564A (en) A method and system for semantic segmentation of small objects in complex scene images
CN113177602B (en) Image classification method, device, electronic equipment and storage medium
CN115661904A (en) Data labeling and domain adaptation model training method, device, equipment and medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant